Michael Perdue Michael Perdue - 1 year ago 101
R Question

R: Diagnostic Tests on Panel Regression

I have the following data that I am running panel regression on. I would like to test for unit root using

plm
's
purtest
:

date years_exp cid cfcontrol leg_totalbills state amtsumlag_1 amtsumfwd_1 billsumfwd_1 billsum amtsum unemplag_1 logleg_totalbills logamtsumlag_1 logamtsumfwd_1 logamtsum exp_5_9 exp_10_19 exp20 post2009years amtsum_5to9exp amtsum_10to19exp amtsum_exp20 amtsumlag1_5to9exp amtsumlag1_10to19exp amtsumlag1_exp20 amtsumfwd1_5to9exp amtsumfwd1_10to19exp amtsumfwd1_exp20 amtsum_post2009years amtsumfwd1_post2009years amtsumlag1_post2009years logamtsum_5to9exp logamtsum_10to19exp logamtsum_exp20 logamtsumlag1_5to9exp logamtsumlag1_10to19exp logamtsumlag1_exp20 logamtsumfwd1_5to9exp logamtsumfwd1_10to19exp logamtsumfwd1_exp20 logamtsum_post2009years logamtsumfwd1_post2009years logamtsumlag1_post2009years
1 2006 18 N00007665 0 2 HI 309 376.54 1 0 286.139 2.9 0.6931472 5.733341 5.931024 5.656478 0 1 0 2006 0 286.139 0 0 309 0 0 376.54 0 573994.8 755339.2 619854 0 5.656478 0 0 5.733341 0 0 5.931024 0 11346.89 11897.63 11501.08
2 2008 20 N00007665 0 2 HI 376.54 13.66 2 0 334.095 2.8 0.6931472 5.931024 2.614472 5.811425 0 0 1 2008 0 0 334.095 0 0 376.54 0 0 13.66 670862.8 27429.28 756092.3 0 0 5.811425 0 0 5.931024 0 0 2.614472 11669.34 5249.859 11909.5
3 2006 24 N00001143 0 2 NY 124.763 140.755 1 0 197.874 5 0.6931472 4.826416 4.947021 5.28763 0 0 1 2006 0 0 197.874 0 0 124.763 0 0 140.755 396935.2 282354.5 250274.6 0 0 5.28763 0 0 4.826416 0 0 4.947021 10606.99 9923.724 9681.79
4 2008 26 N00001143 0 22 NY 140.755 151.65 0 0 243.545 4.6 3.091042 4.947021 5.021575 5.495302 0 0 1 2008 0 0 243.545 0 0 140.755 0 0 151.65 489038.4 304513.2 282636 0 0 5.495302 0 0 4.947021 0 0 5.021575 11034.57 10083.32 9933.618
5 2010 28 N00001143 0 9 NY 151.65 138.25 0 0 194.317 8.3 2.197225 5.021575 4.929064 5.269491 0 0 1 1 0 0 194.317 0 0 151.65 0 0 138.25 194.317 138.25 151.65 0 0 5.269491 0 0 5.021575 0 0 4.929064 5.269491 4.929064 5.021575


I run the following code for unit root using the
hadri, madwu, ips, and levinlin
tests. :

purtest(billsum~trend, data = femodel_1, index=c('date', 'cid'), pmax = 2, exo = "trend", test = "hadri")


and get the following error:

Error in data.frame(
2005
= structure(c(333L, 333L, 55L, 55L, 55L, 55L, : arguments imply differing number of rows: 51, 378, 41, 439, 63, 438, 42, 432, 93, 441


Does anyone have an idea of what might be the problem?

Answer Source

The Hadri LM test for panel unit roots is not applicable to unbalanced panels.

Recommended from our users: Dynamic Network Monitoring from WhatsUp Gold from IPSwitch. Free Download